Speaker: Dr. Norman Gentner
Senior Science Advisor,
Health & Environmental Sciences Division, AECL
Topic: Update on Cancer Mortality from the A-Bombs & Radiation Risk
Location: J.L.Gray Engineering Centre
Deep River, Ont.
Date: Thursday, March 20, 1997 (8:00 pm)

Summary published in North Renfrew Times, March 26, 1997:

What the A-Bombs Tell Us, And What They Don't

by Jeremy Whitlock

"Much ado about ... something very small" could describe the current debate over radiation risk at low exposures, as a capacity crowd learned last Thursday evening at a public seminar of the Canadian Nuclear Society in the J.L.Gray Centre. Dr. Norman Gentner, AECL's Senior Science Advisor for Health and Environmental Sciences, presented an "Update on Cancer Mortality from the A-Bombs and Radiation Risk". The subject, familiar to all those connected with anything nuclear, is a tricky one to publicly communicate: it is fundamental to the widespread public interest in nuclear safety issues, and yet a familiarity with epidemiology and statistical reporting is required for true understanding. The inevitable mismatch of high public concern for low risk is the well-known result.

The impetus for Dr. Gentner's talk was a recent update of the A-bomb Life Span Study (LSS) released by the Radiation Effects Research Foundation. This study, which follows up the survivors of the Hiroshima and Nagasaki bombings, provides the main body of evidence leading to international estimations of radiation risk. Dr. Gentner is especially qualified to assess this recent update: he is an internationally-respected expert on the biological effects of radiation, a member of UNSCEAR (the UN's world-level committee on the subject), and has participated in numerous international projects and working groups. Among these are Chernobyl-related consultations, including his recent role as technical expert to the World Health Organization (WHO), examining the elevated thyroid cancer incidence in Belarus.

The update to the A-bomb LSS adds five more years of observation (1986 to 1990), and 10,536 more survivors, for a total of about 550,000 more person-years of data. This extra statistical power, along with the fact that about 25% of the excess solid cancers occurred in the last five years of the follow-up, makes the latest assessment particularly significant. There have been a total of 421 excess cancer deaths in the survivor group, representing about 8% of the total cancer deaths in the exposed group. From this data, along with the estimations of the dose received in 1945 by each survivor, scientists estimate the dose- response of radiation-induced cancer.

The updated LSS upholds the previous finding of a linear relationship between dose and cancer mortality, with no statistically-significant evidence of nonlinearity in the range 0-3 Sv (0-300 rem). While the LSS, as usual, lacks sufficient accuracy to predict a dose-response at low doses, the improved statistical power of this updated version moves the boundary of significance lower. The "onset" of statistically significant excess risk is now about 50 mSv (5 rem).

At low doses, characteristic of occupational and public exposures, scientists use the linear high-dose response as a prudent guide. A "dose and dose-rate effectiveness factor" (DDREF) of 2 is applied to account for the lower effectiveness in this range. Dr. Gentner cautioned against believing the many statements seen lately claiming "proof" of a zero health effect at low doses. Such studies usually lack statistical power, he says, or are flawed in other ways that aren't immediately apparent to the non-statistician. The "absence of finding" is not the same as the finding of absence.

The relevant discussion, according to Dr. Gentner, involves improvement in the DDREF applied to low doses. In particular, more attention must be paid to the apparent beneficial effect of low dose-rate, as well as dose-fractionation (separation into smaller exposures). He points to a follow-up study of Canadian TB patients receiving fluoroscopy that indicates, in a very strong statistical manner, a zero effect of fairly high doses received in a highly fractionated manner (several days between exposures). Such high-powered studies, enabled by a combination of reasonably high total doses and populations, are the key to understanding the effects of occupational exposures. Additionally, advances in molecular epidemiology and the Human Genome Project will help in the identification of cancer genes.

As an epilogue, Dr. Gentner put the whole matter into perspective by comparing radiation risk with the risk of heart disease, and an apparent large dependence of the latter on socioeconomic position. The quibbling over low-level radiation often obscures the reality that it is indeed a very poor killer of human beings.